Real-Time Classifications of Pests from Agriculture Industry Based on Their Color and Texture Features by Using Machine Learning Models

Abstract

Insects and pests are major sources for causes of agricultural damage. Earlier identification and resistance of pest insects is the only solution to this real-time problem. The combination of computer vision made possible by Machine Learning (ML) has paved the way for smart devices assisted crop insect pest identification. The proposed work aims to develop a classification system for pests using machine learning algorithms. Implementing color and textural features extracted from K-Means clustering and GLCM provides an explicit and comprehensive representation of pest color and texture, enabling the model to classify the pests. Incorporating K-means clustering and GLCM algorithms further enhances the proposed model's ability to differentiate between pests based on their unique features and patterns. This study introduces a novel supervised ML approach for classifying agricultural affecting pests that integrate K-means clustering with color and textural features extracted through the Gray-Level Co-occurrence Matrix GLCM algorithm. Using these advanced techniques, the model achieved an impressive accuracy rate of 98.5% for ten classes of pest classification work. Overall, this study highlights the significance of integrating supervised ML methods with specialized image analysis techniques for precise pest classification in agricultural sectors. The obtained results underscore the potential impact of these advanced technologies on improving pest control and overall agricultural productivity.